Predicting subscriber experience based on QoE

Abstract

The disclosed technology teaches alerting a network operator to deteriorated wireless network conditions with granularity and reduced false alerts, including receiving tracked performance indicators (KPIs) for multiple components of voice, data and messaging and accessing a mapping for combining the performance indicators into quality sub-indicators for individual users and into key quality indicators (KQI) for the individual users. This includes smoothing the current KQI measures of performance over time to reduce the impact of brief transitory events, accessing a weekly profile of KQIs, and comparing KQIs for a current time window to past KQIs for a corresponding time window for the individual users operating mobile devices in part of the cellular network that is a focus of interest. Based on the comparing, the disclosed technology includes generating alerts that report out-of-range current KQIs within the focus of interest that are persistently out-of-range for longer than a configurable time.

We claim as follows: 
     
       1. A method for alerting a network operator to deteriorated wireless network conditions with granularity and reduced false alerts, including:
 receiving tracked performance indicators for multiple components of at least voice, data and messaging over a cellular network; 
 accessing a mapping of the performance indicators to key quality indicators (KQI); 
 combining the performance indicators, according to the mapping, into quality sub-indicators for individual users and into key quality indicators for the individual users; 
 smoothing current KQI measures of performance over time to reduce impact of brief transitory events; 
 accessing a KQI profile of key quality indicators; 
 comparing KQIs for a current time window to past KQIs for a corresponding time window in the KQI profile for the individual users operating mobile devices in part of the cellular network that is a focus of interest; 
 monitoring the individual users by counting instances for each individual user in which the KQIs are out of range for longer than a predetermined time; and 
 based on the comparing, generating alerts that report out-of-range current KQIs within the focus of interest that are persistently out-of-range longer than a configurable time. 
 
     
     
       2. The method of  claim 1 , wherein thousands to millions of users in a cohort are within the focus of interest. 
     
     
       3. The method of  claim 1 , further including contacting the individual user when the counted instances exceed a configurable limit. 
     
     
       4. The method of  claim 1 , further including recording an alert for the individual user when the counted instances exceed a configurable limit. 
     
     
       5. The method of  claim 1 , wherein at least hundreds of users using a particular cell tower are the focus of interest. 
     
     
       6. The method of  claim 1 , further including accessing a mapping of the performance indicators that include a mapping through intermediate quality sub-indicators (SKQI) to KQIs. 
     
     
       7. The method of  claim 1 , further including
 monitoring of cell tower transmission areas as the focus of interest, 
 detecting decreased successful activity level within a cell tower transmission area reflected by the KQIs, and 
 generating an alert when the decreased successful activity level is persistently out-of-range for longer than a configurable predetermined time. 
 
     
     
       8. The method of  claim 1 , further including the KQI profile reflecting an average between corresponding time windows over at least two weeks. 
     
     
       9. The method of  claim 1 , further including the KQI profile reflecting handling missing performance indicators by substituting last average values for the missing performance indicators. 
     
     
       10. The method of  claim 1 , further including handling out-of-range performance indicators that are more than four standard deviations from a mean value for the KQI by substituting values that are in a range of three to four standard deviations from the mean value. 
     
     
       11. A tangible non-transitory computer readable storage media impressed with computer program instructions that, when executed, objectively evaluate customer service experience with data communication services in a wireless communications network, including:
 receiving tracked performance indicators for multiple components of at least voice, data and messaging over a cellular network; 
 accessing a mapping of the performance indicators to key quality indicators (KQI); 
 combining the performance indicators, according to the mapping, into quality sub-indicators for individual users and into KQIs for the individual users; 
 smoothing current KQI measures of performance over time to reduce impact of brief transitory events; 
 accessing a KQI profile of key quality indicators; 
 comparing KQIs for a current time window to past KQIs for a corresponding time window in the KQI profile for the individual users operating mobile devices in part of the cellular network that is a focus of interest; 
 monitoring the individual users by counting instances for each individual user in which the KQIs are out of range for longer than a predetermined time; and 
 based on the comparing, generating alerts that report out-of-range current KQIs within the focus of interest that are persistently out-of-range longer than a configurable time. 
 
     
     
       12. The tangible non-transitory computer readable storage media of  claim 11 , wherein thousands to millions of users in a cohort are within the focus of interest. 
     
     
       13. The tangible non-transitory computer readable storage media of  claim 11 , further including contacting the individual user when the counted instances exceed a configurable limit. 
     
     
       14. The tangible non-transitory computer readable storage media of  claim 11 , further including recording an alert for the individual user when the counted instances exceed a configurable limit. 
     
     
       15. A system for alerting a network operator to deteriorated wireless network conditions with granularity and reduced false alerts includes one or more processors coupled to memory, the memory loaded with computer instructions, that when executed on the processors, implement:
 receiving tracked performance indicators for multiple components of at least voice, data and messaging over a cellular network; 
 accessing a mapping of the performance indicators to key quality indicators (KQI); 
 combining the performance indicators, according to the mapping, into quality sub-indicators for individual users and into key quality indicators for the individual users; 
 smoothing current KQI measures of performance over time to reduce impact of brief transitory events; 
 accessing a KQI profile of key quality indicators; 
 comparing KQIs for a current time window to past KQIs for a corresponding time window in the KQI profile for the individual users operating mobile devices in part of the cellular network that is a focus of interest; 
 monitoring the individual users by counting instances for each individual user in which the KQIs are out of range for longer than a predetermined time; and 
 based on the comparing, generating alerts that report out-of-range current KQIs within the focus of interest that are persistently out-of-range longer than a configurable time. 
 
     
     
       16. The system of  claim 15 , wherein at least hundreds of users using a particular cell tower are the focus of interest. 
     
     
       17. The system of  claim 15 , further including accessing a mapping of the performance indicators that include a mapping through intermediate quality sub-indicators (SKQI) to KQIs. 
     
     
       18. The system of  claim 15 , further including
 monitoring of cell tower transmission areas as the focus of interest, 
 detecting decreased successful activity level within a cell tower transmission area reflected by the KQIs, and 
 generating an alert when the decreased successful activity level is persistently out-of-range for longer than a configurable predetermined time.